Learn R Programming

PopED

PopED computes optimal experimental designs for both population and individual studies based on nonlinear mixed-effect models. Often this is based on a computation of the Fisher Information Matrix (FIM).

Installation

You need to have R installed. Download the latest version of R from www.r-project.org. You can install the released version of PopED from CRAN with:

install.packages("PopED")

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("andrewhooker/PopED")

Getting started

To get started you need to define

A model.
An initial design (and design space if you want to optimize).
The tasks to perform.

Learn more in this introduction to PopED

Contact

You are welcome to:

Ask questions at: https://github.com/andrewhooker/PopED/discussions
submit suggestions and bug-reports at: https://github.com/andrewhooker/PopED/issues
send a pull request on: https://github.com/andrewhooker/PopED
compose a friendly e-mail to: andrew.hooker@farmaci.uu.se

Copy Link

Version

Install

install.packages('PopED')

Monthly Downloads

469

Version

0.7.0

License

LGPL (>= 3)

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/andrewhooker/PopED

Homepage

https://andrewhooker.github.io/PopED/

Maintainer

Andrew C. Hooker

Last Published

October 7th, 2024

Functions in PopED (0.7.0)

Model linearization with respect to epsilon.

Optimize using line search

Optimize the objective function using an adaptive random search algorithm for D-family and E-family designs.

Optimization function for D-family, E-family and Laplace approximated ED designs

PopED - Population (and individual) optimal Experimental Design.

Trace optimization routines

D-family optimization function

The linearized matrix L

Model linearization with respect to epsilon and eta.

Model linearization with respect to occasion variability parameters.

Header function for optimization routines

calc_ofv_and_fim

Calculate the Fisher Information Matrix (FIM) and the OFV(FIM) for either point values or parameters or distributions.

Create a cell array (a matrix of lists)

Build PopED parameter function from a model function

Summarize your optimization settings for optimization routines

calc_ofv_and_grad

Compute an objective function and gradient

Result function for optimization routines

Compute the autofocus portion of the stochastic gradient routine

Nonlinear minimization using BFGS with box constraints

Summarize your experiment for optimization routines

downsizing_general_design

Downsize a general design to a specific design

Compute efficiency.

Create design variables for a full description of a design.

Display a summary of output from poped_db

create.poped.database

Create a PopED database

convert_variables

Create global variables in the PopED database

Evaluate the expectation of the Fisher Information Matrix (FIM) and the expectation of the OFV(FIM).

Evaluate the expectation of determinant the Fisher Information Matrix (FIM) using the Laplace approximation.

Function written to match MATLAB's diag function

create_design_space

Create design variables and a design space for a full description of an optimization problem.

RUV model: Additive .

RUV model: Additive and Proportional.

evaluate_fim_map

Compute the Bayesian Fisher information matrix

evaluate.e.ofv.fim

Evaluate the expectation of the Fisher Information Matrix (FIM) and the expectation of the OFV(FIM).

evaluate_design

Evaluate a design

MATLAB feval function

Evaluate the Fisher Information Matrix (FIM)

RUV model: Proportional.

Power of a design to estimate a parameter.

extract_norm_group_fim

Extract a normalized group FIM

Extract all model parameters from the PopED database.

Compute the expected parameter relative standard errors

ff.PK.1.comp.oral.sd.KE

Structural model: one-compartment, oral absorption, single bolus dose, parameterized using KE.

ff.PK.1.comp.oral.md.KE

Structural model: one-compartment, oral absorption, multiple bolus dose, parameterized using KE.

ff.PKPD.1.comp.sd.CL.emax

Structural model: one-compartment, single bolus IV dose, parameterized using CL driving an EMAX model with a direct effect.

ff.PK.1.comp.oral.sd.CL

Structural model: one-compartment, oral absorption, single bolus dose, parameterized using CL.

ff.PK.1.comp.oral.md.CL

Structural model: one-compartment, oral absorption, multiple bolus dose, parameterized using CL.

ff.PKPD.1.comp.oral.md.CL.imax

Structural model: one-compartment, oral absorption, multiple bolus dose, parameterized using CL driving an inhibitory IMAX model with a direct effect.

MATLAB fileparts function

getTruncatedNormal

Generate a random sample from a truncated normal distribution.

The full Fisher Information Matrix (FIM) for one individual Calculating one model switch at a time, good for large matrices.

median_hilow_poped

Wrap summary functions from Hmisc and ggplot to work with stat_summary in ggplot

Compute the monte-carlo mean of a function

Compute the natural log of the PDF for the parameters in an E-family design

Function written to match MATLAB's isempty function

The Fisher Information Matrix (FIM) for one individual

Model linearization with respect to epsilon.

Compute the inverse of a matrix

get_unfixed_params

Return all the unfixed parameters

Create a full D (between subject variability) matrix given a vector of variances and covariances. Note, this does not test matching vector lengths.

optimize_groupsize

Title Optimize the proportion of individuals in the design groups

Evaluate a criterion of the Fisher Information Matrix (FIM)

model_prediction

Model predictions

Modified Fedorov Exchange Algorithm

Normalize an objective function by the size of the FIM matrix

Create a matrix of ones

Optimize a function using adaptive random search.

Optimize a function using a line search algorithm.

Evaluate the Fisher Information Matrix (FIM)

Translate efficiency to number of subjects

plot_model_prediction

Plot model predictions

Optimize the number of subjects based on desired uncertainty of a parameter.

Run the graphical interface for PopED

Optimization main module for PopED Optimize the objective function. The function works for both discrete and continuous optimization variables. If more than one optimization method is specified then the methods are run in series. If loop_methods=TRUE then the series of optimization methods will be run for iter_max iterations, or until the efficiency of the design after the current series (compared to the start of the series) is less than stop_crit_eff.

Optimize a design defined in a PopED database

Parameter simulation

plot_efficiency_of_windows

Plot the efficiency of windows

Optimization main module for PopED

Choose between arg1 and arg2

Optimization main module for PopED

Timer function (as in MATLAB)

Start parallel computational processes

Timer function (as in MATLAB)

Function written to match MATLAB's randn function

Function written to match MATLAB's size function

Predict shrinkage of empirical Bayes estimates (EBEs) in a population model

summary.poped_optim

Display a summary of output from poped_optim

Retired optimization module for PopED

Test to make sure that matricies are the right size

Function written to match MATLAB's rand function

tryCatch both warnings (with value) and errors

Create a matrix of zeros.